Using Spatial Resampling to Assess Redd Count Survey Length
Requirements for Pacific Lamprey
Mayfield, M. P., Schultz, L. D., Wyss, L. A., Colvin, M. E., & Schreck, C. B. (2014).
Using spatial resampling to assess redd count survey length requirements for
Pacific Lamprey. North American Journal of Fisheries Management, 34(5),
923-931. doi:10.1080/02755947.2014.932867
10.1080/02755947.2014.932867
Taylor & Francis
Version of Record
http://cdss.library.oregonstate.edu/sa-termsofuse
North American Journal of Fisheries Management 34:923–931, 2014
C American Fisheries Society 2014
ISSN: 0275-5947 print / 1548-8675 online
DOI: 10.1080/02755947.2014.932867
ARTICLE
Using Spatial Resampling to Assess Redd Count Survey
Length Requirements for Pacific Lamprey
M. P. Mayfield, L. D. Schultz,* L. A. Wyss,1 and M. E. Colvin
Oregon Cooperative Fish and Wildlife Research Unit, Department of Fisheries and Wildlife,
Oregon State University, 104 Nash Hall, Corvallis, Oregon 97331-3803, USA
Downloaded by [Oregon State University] at 09:29 23 December 2014
C. B. Schreck
U.S. Geological Survey, Oregon Cooperative Fish and Wildlife Research Unit,
Department of Fisheries and Wildlife, Oregon State University, 104 Nash Hall, Corvallis,
Oregon 97331-3803, USA
Abstract
Pacific Lamprey Entosphenus tridentatus has declined across its range along the West Coast of North America,
and an understanding of all life history phases is needed to address population recovery. Spawning surveys (redd
counts) are common tools currently used to monitor returning adult salmonids, but such methodologies are in their
infancy for Pacific Lamprey. Our objective was to assess the minimum spawning survey distance required to detect
the presence of Pacific Lamprey redds and obtain precise redd density estimates from these data. To do this, we
statistically resampled existing spawning locations of Pacific Lamprey collected during spawning surveys in four
streams of the Willamette River Basin, Oregon, during spring of 2013. We found that the minimum survey distance
for Pacific Lamprey redd detection was inversely related to the observed redd density and was always less than 1.2 km.
Survey distance requirements to obtain precise redd counts ( ± 20% of observed redd densities) were also inversely
related to redd density and habitat availability, and varied between 1.3 km and 13.7 km. Our results suggest that
spawning surveys are a potential tool for monitoring adult Pacific Lamprey abundance, but the specific objectives
of the monitoring programs and acknowledgment of unknowns must be considered prior to implementation into
recovery plans.
Pacific Lamprey Entosphenus tridentatus is considered a cultural and ecological icon for native peoples of the Pacific Northwest (CRITFC 2011). The species has declined considerably
across its range in western North America, and complete extirpation has occurred in multiple drainages (Moser and Close 2003).
Between 1939 and 1969, Pacific Lamprey counts at Bonneville
Dam on the Columbia River averaged over 100,000 adults but
had declined to an average of less than 40,000 between 1997
and 2010 (Murauskas et al. 2013). The Columbia River is not
the only location that has experienced a decline of Pacific Lamprey runs; annual abundance in 2009 and 2010 were estimated
at less than 500 individuals at the Winchester Dam on the North
Umpqua River, while historical adult abundance (1965–1971)
was estimated at between 14,532 and 46,785 (Goodman et al.
2005). Despite the many important ecological and cultural roles
of Pacific Lamprey (outlined in Close et al. 2002), conservation
efforts have been hampered by numerous knowledge gaps in its
biology (Luzier et al. 2011). Of the areas where the species has
persisted, the Willamette River, Oregon, has one of the highest remaining adult abundances (Kostow 2002), and Willamette
Falls (approximately 42 km upstream of the Columbia River
confluence) currently supports one of the few remaining harvest
locations for Pacific Lamprey. Although migrating adult populations can be monitored at natural and man-made barriers, little
*Corresponding author: luke.schultz@oregonstate.edu
1
Present address: Calapooia, North Santiam, and South Santiam Watershed Councils, Post Office Box 844, Brownsville, Oregon 97327, USA.
Received January 2, 2014; accepted May 19, 2014
923
Downloaded by [Oregon State University] at 09:29 23 December 2014
924
MAYFIELD ET AL.
is known about Pacific Lamprey spawning distribution, characteristics (e.g., abundance, timing, habitat use), or populations
within tributaries (CRITFC 2011).
To understand how Pacific Lamprey use available habitat
in a large river system, it is important to develop methods for
monitoring all life history stages, including the spawning stage.
Redd surveys have been used for many years to monitor both
anadromous and nonanadromous salmonid populations due to
their nonintrusive nature, relatively low cost, and ability to survey large areas with little effort (Gallagher et al. 2007; Dunham
et al. 2009). While redd count methods have been routinely
implemented for salmonids, the development of standardized
spawning surveys for Pacific Lamprey is in its infancy (Stone
2006; Brumo et al. 2009; Gunckel et al. 2009). In general,
salmonid and Pacific Lamprey redds can be counted by using
similar methods, but the assumptions related to redd counts,
specifically redd detection rates, must be examined and addressed for Pacific Lamprey (Dunham et al. 2009).
Spawning surveys for Pacific Lamprey have traditionally
been problematic for several reasons, including confusion of
Pacific Lamprey redds with salmonid redds (Stone 2006), false
identification in low- or high-density areas (Brumo et al. 2009),
questions about redd longevity and loss (Stone 2006), and the
relationship between redd counts and spawning adult numbers
(Moser et al. 2007). Despite the potential sources of error, redd
counts remain an efficient method to monitor populations and
to assess Pacific Lamprey distribution, particularly with proper
surveyor training (Wyss et al. 2013). For redd counts to be an
effective monitoring tool, it will be important to ensure that
surveys are conducted in a manner that both reduces bias and
increases efficiency. Our objectives were to evaluate the required length (i.e., distance) of redd surveys for spawning Pacific Lamprey to allow for redd detection and to estimate redd
density within an acceptable level of precision. To address this
objective, we spatially resampled georeferenced spawning locations of Pacific Lamprey. By determining the suitable distance
for surveying Pacific Lamprey redds, our analyses will increase
surveying efficiency, allow sampling of larger spatial areas, and
help address pertinent management questions.
METHODS
Redd counts.—We monitored Pacific Lamprey spawning activity from late April through mid-June 2013 in four tributaries
to the Willamette River Basin: Calapooia River, Luckiamute
River, Marys River, and Thomas Creek (Figure 1). Our spawning survey methodologies were based on those used by other
adult Pacific Lamprey studies (Stone 2006; Brumo et al. 2009;
Gunckel et al. 2009) and included insights from preliminary
spawning surveys conducted in the Willamette River Basin during 2011 (Clemens et al. 2012) and from redd counts for anadromous salmonid species (Gallagher et al. 2007). Survey locations were selected based on stream access, our ability to safely
float in personal watercraft, and streams where larval Pacific
Lamprey sampling was also conducted (Wyss et al. 2012,
2013). We selected survey segments to represent available habitat in each tributary; survey segments included all underlying
geologic types present in each tributary (see below).
We conducted spawning surveys in multiple tributaries to
the Willamette River to capture the spatial variability of spawning activity. Redd count surveys were initiated during 2012 to
evaluate and refine survey methodology. These included two
survey segments on both the Marys River and Thomas Creek;
the Marys River had contiguous survey segments, whereas
the survey segments on Thomas Creek were spatially discrete
(Figure 1). In 2013, redd count surveys continued in those
streams and we added single survey segments on the Calapooia
and Luckiamute rivers (Figure 1). We had no a priori knowledge
of the current distribution of Pacific Lamprey in these latter two
rivers, except for radiotelemetric indications that adults had entered these streams near their confluence with the Willamette
River (Clemens et al. 2012). We used only 2013 data for the
analysis presented. In that year we surveyed these four streams
with different habitats and peak redd densities, and all surveys
were conducted with the same personnel that were trained to
classify Pacific Lamprey redds. Additionally, relatively low water discharge and increased water clarity in 2013 increased our
confidence in redd observation and counts.
All surveys were conducted by two surveyors in individual
inflatable pontoon boats with each covering an area from stream
bank to midchannel to observe the entire channel. We enumerated all Pacific Lamprey redds, live adults, and carcasses that
were observed in each surveyed segment. Although subjectivity
exists in redd surveys (Dunham et al. 2001), we defined redds
as round depressions in the stream substrate that appeared to
be actively excavated (Stone 2006). Larger rocks are frequently
moved by Pacific Lamprey to the edge of redd depressions
(Gunckel et al. 2009), an additional characteristic we used to
differentiate Pacific Lamprey redds from those of other fishes.
Additionally, the presence of actively spawning adult Pacific
Lamprey in redds helped confirm our redd identification. Brook
Lamprey Lampetra richardsoni (or L. pacifica) also spawn during the timeframe of Pacific Lamprey. However, median Brook
Lamprey redd size (0.03 m2) is much smaller than median Pacific Lamprey redd size (0.124 m2; Stone 2006; Gunckel et al.
2009). If large areas of substrate appeared to be disturbed or
redds appeared to overlap one another, as has been observed by
Stone (2006) and Brumo et al. (2009), we counted each discrete
depression as an individual redd.
Each segment was visited multiple times from April through
June to capture temporal variability in spawning activity and
to observe peak redd counts (i.e., the highest number of redds
observed in all visits). Because no information is currently available related to the frequency of surveys needed for monitoring
Pacific Lamprey redds, we attempted to visit each survey segment every 10–14 d, as is recommended in the salmonid literature (Gallagher et al. 2007). We considered our sampling recurrence appropriate to detect redds of various ages because median
Downloaded by [Oregon State University] at 09:29 23 December 2014
USING SPATIAL RESAMPLING TO ASSESS REDD COUNTS
925
FIGURE 1. Pacific Lamprey spawning survey locations in four tributaries of the Willamette River, 2013. Segments of streams surveyed are in bold. Additional
details on the distribution of redds within survey segments are shown in Figure 2.
longevity of Pacific Lamprey redds has been reported at 40 d
with a minimum longevity of 10 d (Stone 2006). Occasionally,
weather events and hydrologic conditions precluded completion of our surveys on schedule; those surveys were completed
as soon as stream conditions permitted surveying.
We georeferenced all Pacific Lamprey redds identified during surveys using a hand-held global positioning system unit
to generate redd distribution maps for each survey (Figure 2).
We used linear referencing tools in ArcGIS version 10.1 (ESRI
2012) to determine the stream kilometer (as measured from the
confluence of the tributary with a major river) of each redd location within a survey. We used existing geologic maps (Walker
and MacLeod 1991) to calculate the percent of the stream survey
distance with alluvial underlying geology in each survey segment. Alluvial geology, defined as underlying sediments that
were deposited by moving water, included fine-grain alluvial,
medium-grain alluvial, and Missoula Flood alluvial geologic
types from geologic maps. In our related work (Schultz et al.
2014), we found that Pacific Lamprey selected for reaches with
alluvial geology across the Willamette River Basin, and the
resulting spatial distribution of redds (i.e., “clumpiness”) was
strongly related to these underlying geologic patterns. Our surveys encompassed a wide range of stream types and habitats
on both the west and east sides of the Willamette Basin, with
headwaters located in the Coast Range and Cascade Mountains,
respectively (Table 1). We conducted several surveys in each
segment at different times during the spawning season, but the
peak redd density observed for each survey segment was used
for analysis purposes, as is commonly done with redd count data
analyses (Gallagher et al. 2007).
Spatial resampling analysis.—We used spatial resampling of
the georeferenced redd locations of each survey segment to evaluate suitable survey lengths (i.e., distance) in future sampling.
A sequence of simulated subsample lengths was developed, beginning with 0.1 km and continuing by adding 0.1 km to every
subsequent survey segment length until reaching a final subsample length of 1.0 km less than the total survey segment length
(e.g., the Calapooia River survey was 7.3 km, so the subsample
lengths we used varied from 0.1 to 6.3 km at 0.1-km intervals).
For each subsample length, a continuous segment of that length
was randomly selected from within our spatial data set of redd
locations and the density of redds in that segment was calculated.
MAYFIELD ET AL.
Downloaded by [Oregon State University] at 09:29 23 December 2014
926
FIGURE 2. Distribution of Pacific Lamprey redds in the upper survey segment of Thomas Creek and in the Marys River, Oregon. For each location where redds
were observed, the number of redds at that location is indicated by the size of the symbol, and the map indicates the degree of clumpiness of Pacific Lamprey redds
for relatively high and low redd densities in upper Thomas Creek and Marys River, respectively. We spatially resampled from this observed distribution to assess
minimum survey distance to detect redds and obtain redd densities within 20% of observed. Note: inset is Figure 1.
This procedure was repeated (with replacement) 10,000 times
for each subsample length. Subsamples were restricted to be
fully within the actual surveyed stream segment to ensure that
no edge correction was needed. From each simulated subsample
length, we determined the mean, maximum, and minimum redd
density values from the 10,000 resamples. All resampling simulations were completed in Program R (R Core Development
Team 2013).
927
USING SPATIAL RESAMPLING TO ASSESS REDD COUNTS
TABLE 1. Observed peak redd density, percent alluvial geologic composition, and survey length requirements of three spawning activity metrics for survey
segments in the Willamette River Basin. Survey length requirements were calculated from spatial resampling simulations to determine length needed to detect
Pacific Lamprey spawning presence (>0.0 redds/km) and redd densities within acceptable error range (lower error level = 80% of observed, upper error level =
120% of observed). Mean stream width was measured during summer base-flow conditions.
Survey length requirements (km)
Survey
Downloaded by [Oregon State University] at 09:29 23 December 2014
Marys River
Luckiamute River
Calapooia River
Thomas Creek, upper
Thomas Creek, lower
Segment
length (km)
Observed peak
(redds/km)
Mean stream
width (m)
% Alluvial
Detection
80% of
observed
120% of
observed
19.2
5.6
7.3
4.2
7.0
22.9
47.2
88.5
116.2
137.0
17.3
16.8
15.7
19.6
19.7
67
80
100
100
100
1.2
0.7
0.5
0.3
0.3
10.4
4.9
2.8
2.4
2.4
13.7
2.4
3.0
1.3
3.2
We used the mean, maximum, and minimum redd density
values for each subsample to determine both the minimum survey distance necessary to detect Pacific Lamprey redds and the
distance required to obtain precise redd density estimates. The
minimum distance to detect redds was determined by identifying the survey length in which all simulations had a redd
density greater than 0.0 redds/km. We considered redd density
estimates to be precise if both the minimum and maximum redd
counts for each simulated segment length were within ± 20%
of the observed redd density for the entire survey segment. We
included this measure of error to account for surveyor-related
differences in redd detection rates. Our experience with Pacific
Lamprey redd surveys provided evidence that our redd counts
were minimally biased, but it is likely that we and others might
misidentify or omit redds during surveys. Although there is
much variation in redd detection rates in the literature, we chose
the 20% error based on the estimated detection rates of 0.83
from surveys of bull trout redds (Muhlfeld et al. 2006), but were
slightly more conservative by rounding this to an estimated detection rate of 0.80. This estimate is conservative based on other
salmonid redd counts, in which estimated individual detection
rates ranged from 0.28 to 2.54 (Dunham et al. 2001; Gallagher
and Gallagher 2005).
RESULTS
Redd Count Surveys
During the 2013 spawning season, we surveyed 43.3 km
of stream habitat in locations throughout the Willamette River
Basin, with individual survey lengths ranging from 4.2 km in
upper Thomas Creek to 19.2 km of total continuous survey
distance in Marys River (Table 1). Peak redd densities differed among streams, varying from 22.9 redds/km in the Marys
River to 137.0 redds/km in the lower segment of Thomas Creek
(Table 1). Although our sample size and the range of observed
percentage of alluvial underlying geology were limited, variation in redd densities appeared to be related to the percentage
of alluvial underlying geology; streams with lower observed
redd densities (i.e., the Marys and Luckiamute rivers) had a
lower proportion of alluvial geology, while the Calapooia River
and Thomas Creek consisted entirely of alluvial geology and
had much higher redd densities (Table 1). In the two streams
with multiple years of survey data (i.e., the Marys River and
Thomas Creek), the redd densities from 2013 were similar to
those observed in previous years.
Spatial Resampling Simulations
We observed an inverse relationship between the minimum
survey distance needed to detect Pacific Lamprey redds and
observed redd density with our spatial resampling simulations.
In the Marys River, the survey with the lowest redd density, a
survey distance of 1.2 km was needed to detect redds (Table 1
and Figure 3). In contrast, in both sections of Thomas Creek,
which had our highest observed redd densities, only 0.3 km
was needed to detect redds (Table 1 and Figure 4). For the
Luckiamute and Calapooia rivers, two streams with moderate
redd densities, the survey lengths needed for detection were 0.7
and 0.5 km, respectively (Table 1 and Figures 3 and 4).
The survey length needed to obtain simulated redd densities
within ± 20% of observed redd densities generally decreased
as observed redd density and the percentage of alluvial geology increased (Table 1). Because the minimum and maximum
simulated redd density values were not symmetrical around the
mean, the required survey lengths for each stream differed for
the lower error level (80% of observed redd density) and the
upper error level (120% of observed redd density); in our assessments, we used the longer survey length. The three surveys
with the highest redd densities (upper and lower Thomas Creek
and the Calapooia River) required minimum survey lengths of
2.4, 3.2, and 3.0 km, respectively, to obtain a subsample within
± 20% of the observed redd density (Table 1 and Figure 4). In
the streams with lower redd densities, a longer survey length
was needed to obtain precise redd density estimates. For the
Luckiamute River, a survey length of 4.9 km was needed to
obtain precise redd density estimates, whereas the Marys River
required a survey length of 13.7 km.
Downloaded by [Oregon State University] at 09:29 23 December 2014
928
MAYFIELD ET AL.
FIGURE 3. Spatial resampling simulation results of mean redd density (solid
line), minimum and maximum values (dotted lines), and the observed mean
redd density (dashed line) for study surveys in the Luckiamute and Marys
Rivers, two streams with a mix of alluvial and nonalluvial underlying geology.
Dashed gray lines indicate ± 20% of the observed redd density, and the points
where minimum and maximum values both lie within the dashed lines indicate
the survey length requirements for obtaining a relatively precise redd density
measurement. Note differences in scale of y-axes between panels.
DISCUSSION
Our spawning survey efforts have incorporated methods and
findings from previous Pacific Lamprey spawning surveys in
the Pacific Northwest (Stone 2006; Brumo et al. 2009; Gunckel
et al. 2009) to learn more about monitoring the spawning phase
of the Pacific Lamprey life cycle. Our findings suggest survey
length requirements for detecting and evaluating the abundance
of spawning adult Pacific Lamprey and can be used in the development of monitoring programs. Furthermore, these data may
be used to assess the relative precision of spawning surveys
that have already been completed and adapt survey methodology and biological interpretations accordingly. Our spatial
resampling approach may also be applied to existing data sets
from spawning surveys for other fishes to optimize fieldwork
efficiency. Although redd surveys are generally considered an
imprecise measure of Pacific Lamprey abundance (Moser et al.
2007), redd counts may facilitate trend assessments of status
more accurately than have traditional measures of adult abundance such as counts at dams, which are often poor reflections
of escapement (Moser and Close 2003; Lampman 2011). However, addressing several key uncertainties might allow using
redd counts to compute population metrics for Pacific Lamprey.
Although we only used survey data from 1 year, which had
relatively good spawning survey conditions, the stream segments we surveyed had a range of redd densities and geologic
composition, and our results indicate how survey length relates
to these factors. Except for the Luckiamute River, our resampled mean redd density showed very little bias relative to what
we observed in our spawning surveys. In the Luckiamute River
survey, however, there was an aggregation of several redds right
FIGURE 4. Spatial resampling simulation results of mean redd density (solid
line), minimum and maximum values (dotted lines), and the observed mean redd
density (dashed line) for study surveys in Thomas Creek and the Calapooia River,
two streams consisting of entirely alluvial underlying geology. Dashed gray lines
show ± 20% of the observed redd density, and the points where minimum and
maximum values both lie within the dashed lines indicate the survey length
requirements for obtaining a relatively precise redd density measurement. Note
differences in scale of y-axes between panels.
at the end of the survey segment. The negative bias that we
observed in that survey was probably because these redds were
often not included in resamples. Our use of the minimum and
maximum observed values (of 10,000 observations) rather than
other statistics of variance (e.g., 95% confidence intervals) from
our resampling techniques provided the most conservative survey length recommendations possible from our data. Because
of our approach, our results can be applied to inform survey
methods in other basins with a mix of hydrologic conditions,
alluvial and intrusive geology, stream flow regimes, and anthropogenic impacts. However, in such other basins, similar techniques should be used to validate spawning survey lengths and
refine monitoring methodologies.
Implementation of spawning surveys for Pacific Lamprey can
be informed by our analyses, but clearly outlining the objectives
of a monitoring plan for Pacific Lamprey spawning activity will
dictate the logistics of conducting surveys. If detection of Pacific
Lamprey is the study goal, our spatial resampling analyses suggest that completing short spawning surveys (of at least 1.2 km)
over a large spatial scale may be the most effective way to
Downloaded by [Oregon State University] at 09:29 23 December 2014
USING SPATIAL RESAMPLING TO ASSESS REDD COUNTS
meet objectives. If monitoring redd density, temporal trends,
and other spawning characteristics is desired, our results indicate that available spawning habitat and underlying geologic
structure are important to consider to obtain relatively precise
estimates of redd density. Although our survey segments contained a relatively narrow range of percent alluvial underlying
geology, some of our related work suggests that Pacific Lamprey select for areas with these underlying sediments (Schultz
et al. 2014). Stream segments with alluvial underlying geology,
in addition to providing a source of suitable spawning gravel,
are also more likely to retain these sediments than are areas with
bedrock streambeds (May and Gresswell 2003), particularly in
streams that have experienced historic habitat degradation (e.g.,
splash dams; Miller 2010) and loss of large wood (Sedell and
Froggatt 1984). In other areas, particularly river basins with little
alluvial geology, examining the relationship between underlying geology and redd density will be important for developing
and refining spawning survey methodology.
An additional consideration prior to implementing monitoring programs is the site-specific statistical power of redd counts
for population monitoring. By acknowledging a potential error
rate of ± 20% in redd counts, the statistical ability to detect
small temporal changes in populations based on spawning activity is greatly reduced and may not even be possible (Maxwell
1999; Wagner et al. 2013). If greater statistical power is desired
to be able to detect changes in spawning populations, it will
be necessary to increase survey length to capture a larger sample of spawning activity and increase precision of redd counts.
The spatial pattern of Pacific Lamprey redds is often “clumped”
and can deviate significantly from a completely random spatial
pattern (Schultz et al. 2014); as a result, some reaches consist
of short segments with many redds tightly packed in gravel of
suitable size for spawning as well as long segments with no
redds. Even in stream segments with relatively high redd densities, longer (i.e., distance) surveys may be needed to account
for this habitat patchiness. Because redds are often clumped
and the relationship between redd density and adult abundance
is uncertain (Schultz et al. 2014), it may be more effective to
use approaches that account for detection probability and abundance (e.g., a multistate occupancy model; Nichols et al. 2007)
to evaluate trends in adult populations.
Using redd counts to monitor temporal trends in populations
with any amount of statistical power requires the reduction of
bias, particularly bias associated with redd detection (Maxwell
1999; Wagner et al. 2013). Proper training of personnel conducting spawning surveys is essential to ensure the most accurate redd counts possible (Dunham et al. 2009). Our spawning
surveys were conducted by two individuals; one person had
extensive salmonid spawning survey experience and the other
had 1 year of Pacific Lamprey spawning experience. While Pacific Lamprey redds are often located in similar habitat types
(i.e., pool tailouts, riffles) as salmonids, they generally tend
to be much smaller than anadromous salmonids (Stone 2006;
Gunckel et al. 2009; Clemens et al. 2012; Wyss et al. 2013);
929
moreover, Pacific Lamprey typically place excavated substrate
pieces at the upstream end of redds (Stone 2006). Redd superimposition is an additional issue that we and other researchers
(e.g., Al-Chokhachy et al. 2005; Brumo et al. 2009; Gunckel
et al. 2009) have observed during redd surveys and might lead
to an underestimation of redd counts. False identification of
redds may also be more likely to occur in areas with low redd
densities (Muhlfeld et al. 2006). Clearly specifying a priori what
characteristics constitute a counted redd will greatly increase the
precision of redd counts between observers and time conditions
(Wyss et al. 2013). Environmental conditions can also affect
the ability to accurately detect redds; in wet years, it is more
likely that redds will be washed out between surveys (Jones
2012) or that low visibility levels will reduce probabilities of
redd detection.
Pacific Lamprey population metrics and trend assessments
derived from spawning surveys should also be interpreted with
caution (sensu Maxwell 1999). For spawning surveys to provide
cost-effective measures of redd density across stream segments,
the relationship between redds counts and adult abundance, factors that influence variability between surveyors, and male-tofemale sex ratios must be examined (Dunham et al. 2001, 2009;
Parsons and Skalski 2009). While some assumptions of redd
identification error can be obtained from salmonid studies, many
factors are unknown for Pacific Lamprey, including how many
redds are constructed per adult and the proportion of test redds
constructed. Stone (2006) observed one female build multiple
redds with several males, a characteristic that we frequently
observed during our surveys. These types of spawning characteristics must be further investigated before Pacific Lamprey
redd counts can be used for monitoring with more statistical
power.
Despite the unknowns in Pacific Lamprey spawning behavior, redd counts remain useful as a means of understanding
the distribution of this anadromous fish across its range and
can easily be integrated with existing spring spawning surveys.
Pacific Lamprey redds are often encountered during steelhead
Oncorhynchus mykiss spawning surveys, but the surveys usually end before peak Pacific Lamprey spawning has occurred (E.
Brown, Oregon Department of Fish and Wildlife, Oregon Adult
Salmonid Inventory and Sampling Project, personal communication). Extending these steelhead surveys through June would
probably capture the majority of the Pacific Lamprey spawning
season. Logistical constraints would dictate the feasibility of
walking or floating surveys, but we have documented spawning
activities in rivers as small as 3–5 m wide (base flow) to as
large as 100–150 m wide (Schultz et al. 2014). However, the
ability to accurately observe all available spawning habitat in
wide stream segments (>25 m wide) is greatly compromised
relative to smaller habitats. Redd counts also have considerable
potential value as a coarse measure of adult Pacific Lamprey
abundance. Our results provide a starting point for developing a
monitoring design using redd counts for Pacific Lamprey populations. These monitoring plans will provide an informative tool
930
MAYFIELD ET AL.
Downloaded by [Oregon State University] at 09:29 23 December 2014
for monitoring the response of the spawning life history stage
to management actions and help address the ongoing decline of
Pacific Lamprey.
ACKNOWLEDGMENTS
Funding for this study was provided by the Columbia River
Inter-Tribal Fish Commission through the Columbia Basin Fish
Accords partnership with the Bonneville Power Administration
under project 2008-524-00, Brian McIlraith, project manager.
Fieldwork assistance was generously provided by K. McDonnell and C. Christianson. B. McIlraith, B. Heinith, J. Lemanski,
J. Peterson, B. Gerth, G. Giannico, J. Jolley, G. Silver, and R.
Wildman provided sampling and analyses advice and lent field
equipment to our project. A critical review by B. Clemens, R.
Al-Chokhachy, and two anonymous reviewers greatly improved
the quality of this manuscript. Any use of trade, firm, or product names is for descriptive purposes only and does not imply
endorsement by the U.S. Government.
REFERENCES
Al-Chokhachy, R., P. Budy, and H. Schaller. 2005. Understanding the significance of redd counts: a comparison between two methods for estimating the
abundance of and monitoring bull trout populations. North American Journal
of Fisheries Management 25:1505–1512.
Brumo, A. F., L. Grandmontagne, S. N. Namitz, and D. F. Markle. 2009. Approaches for monitoring Pacific Lamprey spawning populations in a coastal
Oregon stream. Pages 203–222 in L. R. Brown, S. D. Chase, M. G. Mesa, R.
J. Beamish, and P. B. Moyle, editors. Biology, management, and conservation
of lampreys in North America. American Fisheries Society, Symposium 72,
Bethesda, Maryland.
Clemens, B. J., L. Wyss, R. McCoun, L. Schwabe, I. Courter, S. Duery, J.
Vaughn, and C. B. Schreck. 2012. Migration characteristics and habitat use
of the imperiled adult Pacific Lamprey in the Willamette Basin: prelude to
estimating requirements for persistence. Annual Report to the Columbia River
Inter-tribal Fish Commission, Portland, Oregon.
Close D., K. Aronsuu, A. Jackson, T. Robinson, J. Bayer, J. Seelye, S. Yun, A.
Scott, W. Li, and C. Torgersen. 2002. Pacific Lamprey research and restoration
project. Annual Report to the Bonneville Power Administration, Project 199402600, Portland, Oregon.
CRITFC (Columbia River Inter-Tribal Fish Commission). 2011. Tribal Pacific Lamprey restoration plan for the Columbia River basin. Available:
http://www.critfc.org/lamprey/lamprey plan.pdf. (January 2012).
Dunham, J., B. Reiman, and K. Davis. 2001. Sources and magnitude of sampling
error in redd counts for Bull Trout. North American Journal of Fisheries
Management 21:343–352.
Dunham, J. B., A. E. Rosenberger, R. F. Thurow, C. A. Dollof, and P. J. Howell.
2009. Coldwater fish in wadeable streams. Pages 119–138 in S. A. Bonar,
W. A. Hubert, and D. W. Willis, editors. Standard methods for sampling
North American freshwater fishes. American Fisheries Society, Bethesda,
Maryland.
ESRI (Environmental Systems Research Institute). 2012. ArcGIS version 10.0.
ESRI, Redlands, California.
Gallagher, S. P., and C. M. Gallagher. 2005. Discrimination of Chinook
Salmon, Coho Salmon, and steelhead redds and evaluation of the use
of redd data for estimating escapement in several unregulated streams in
Northern California. North American Journal of Fisheries Management 25:
284–300.
Gallagher, S. P., P. K. J. Hahn, and D. H. Johnson. 2007. Redd counts.
Pages 197–234 in D. H. Johnson, B. M. Shrier, J. S. O’Neal, J. A.
Knutszen, X. Augerot, T. A. O’Neil, and T. N. Pearsons. Salmonid
field protocols handbook: techniques for assessing status and trends in
salmon and trout populations. American Fisheries Society, Bethesda,
Maryland.
Goodman, K., N. Ackerman, S. Gunckel, R. Beamesderfer, L. Krentz, P. Sheerer,
C. Kern, D. Ward, and C. Ackerman. 2005. Oregon native fish status report.
Oregon Department of Fish and Wildlife, Salem.
Gunckel, S. L., K. K. Jones, and S. E. Jacobs. 2009. Spawning distribution
and habitat use of adult Pacific and Western Brook lampreys in Smith River,
Oregon. Pages 173–189 in L. R. Brown, S. D. Chase, M. G. Mesa, R. J.
Beamish, and P. B. Moyle, editors. Biology, management, and conservation
of lampreys in North America. American Fisheries Society, Symposium 72,
Bethesda, Maryland.
Jones, E. C. 2012. Environmental stochasticity and the reliability of redd count
data: a simulation study of redd construction, redd loss, and survey frequency
in a small coastal stream in Northern California. Master’s thesis. Humboldt
State University, Arcata, California.
Kostow, K. 2002. Oregon lampreys: natural history status and problems analysis.
Oregon Department of Fisheries and Wildlife, Salem.
Lampman, R. T. 2011 Passage, migration behavior, and autoecology of
adult Pacific Lamprey at Winchester Dam and within the North Umpqua
River basin, Oregon, USA. Master’s thesis. Oregon State University,
Corvallis.
Luzier, C. W, H. A. Schaller, J. K. Bostrom, C. Cook-Tabor, D. H. Goodman,
R. D. Nelle, K. Ostrand, and B. Strief. 2011. Pacific Lamprey (Entosphenus
tridentatus) assessment and template for conservation measures. U.S. Fish
and Wildlife Service, Portland, Oregon.
Maxwell, B. A. 1999. A power analysis on the monitoring of Bull Trout stocks
using redd counts. North American Journal of Fisheries Management 19:
860–866.
May, C. L., and R. E. Gresswell. 2003. Processes and rates of sediment and
wood accumulation in headwater streams of the Oregon Coast Range, USA.
Earth Surface Processes and Landforms 28:409–424.
Miller, R. R. 2010. Is the past present? Historical splash-dam mapping and
stream disturbance detection in the Oregon Coastal Province. Master’s thesis.
Oregon State University, Corvallis.
Moser, M. L., J. M. Butzerin, and D. B. Dey. 2007. Capture and collection of
lampreys: the state of the science. Reviews in Fish Biology and Fisheries
17:45–56.
Moser, M. L., and D. A. Close. 2003. Assessing Pacific Lamprey status in the
Columbia River basin. Northwest Science 77:116–125.
Muhlfeld C. C., M. L. Taper, D. F. Staples, and B. B. Shepard. 2006. Observer
error structure in Bull Trout redd counts in Montana streams: implications
for inference on true redd numbers. Transactions of the American Fisheries
Society 135:643–654.
Murauskas, J. G., A. M. Orlov, and K. A. Siwicke. 2013. Relationships between
the abundance of Pacific Lamprey in the Columbia River and their common
hosts in the marine environment. Transactions of the American Fisheries
Society 142:143–155.
Nichols, J. D., J. E. Hines, D. I. Mackenzie, M. E. Seamans, and R. J. Gutiérrez.
2007. Occupancy estimation and modeling with multiple states and state
uncertainty. Ecology 88:1395–1400.
Parsons, A. L. and J. R. Skalski. 2009. The design and analysis of salmonid
tagging studies in the Columbia Basin, volume XXIV. A statistical critique
of estimating salmon escapement in the Pacific Northwest. Report to the
Bonneville Power Administration, Project 1989-107-00, Portland, Oregon.
R Core Development Team. 2013. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna. Available:
http://www.R-project.org./ (July 2014).
Schultz, L. D., M. P. Mayfield, L. A. Wyss, G. T. Sheoships, B. J. Clemens, B.
Chasco, and C. B. Schreck. 2014. The distribution and relative abundance
of spawning and larval Pacific Lamprey in the Willamette River basin. Final Report to the Columbia River Inter-Tribal Fish Commission, Portland,
Oregon.
Sedell J. R., and J. L. Froggatt. 1984. Importance of streamside forests to
large rivers: the isolation of the Willamette River, Oregon, U.S. from its
USING SPATIAL RESAMPLING TO ASSESS REDD COUNTS
Downloaded by [Oregon State University] at 09:29 23 December 2014
flood plain by snagging and streamside forest removal. Internationale Vereinigung fuer Theoretische und Angewandte Limnologie Verhandlungen 22:
1828–1834.
Stone, J. 2006. Observation on nest characteristics, spawning habitat, and spawning behavior of Pacific and Western Brook lamprey in a Washington stream.
Northwestern Naturalist 87:225–232.
Wagner, T., B. J. Irwin, J. R. Bence, and D. B. Hayes. 2013. Detecting temporal
trends in freshwater fisheries surveys: statistical power and the important
linkages between management questions and monitoring objectives. Fisheries
38:309–319.
931
Walker, G. W., and N. S. MacLeod. 1991. Geological map of Oregon. U.S.
Geological Survey, Reston, Virginia.
Wyss, L. A., B. J. Clemens, and C. B. Schreck. 2012. Relative abundance
and associated habitat characteristics of larval lamprey in five Willamette
River tributaries. Annual Report to the Columbia River Inter-Tribal Fish
Commission, Portland, Oregon.
Wyss, L. A., L. Schultz, G. T. Sheoships, and C. B. Schreck. 2013. Monitoring
Pacific Lamprey relative abundance and distribution in the Willamette River
basin. Annual Report to the Columbia River Inter-Tribal Fish Commission,
Portland, Oregon.